Kerple: Kernelized relative positional embedding for length extrapolation TC Chi, TH Fan, PJ Ramadge, A Rudnicky Advances in Neural Information Processing Systems 35, 8386-8399, 2022 | 24 | 2022 |
Dissecting transformer length extrapolation via the lens of receptive field analysis TC Chi, TH Fan, AI Rudnicky, PJ Ramadge Proceedings of the 61st Annual Meeting of the Association for Computational …, 2023 | 20* | 2023 |
Powergym: A reinforcement learning environment for volt-var control in power distribution systems TH Fan, XY Lee, Y Wang Learning for Dynamics and Control Conference, 21-33, 2022 | 15 | 2022 |
Model imitation for model-based reinforcement learning YH Wu, TH Fan, PJ Ramadge, H Su arXiv preprint arXiv:1909.11821, 2019 | 12 | 2019 |
Soft actor-critic with integer actions TH Fan, Y Wang 2022 American Control Conference (ACC), 2611-2616, 2022 | 10 | 2022 |
Training discrete deep generative models via gapped straight-through estimator TH Fan, TC Chi, AI Rudnicky, PJ Ramadge International Conference on Machine Learning, 6059-6073, 2022 | 6 | 2022 |
Rumor source detection: a probabilistic perspective TH Fan, IH Wang 2018 IEEE International conference on acoustics, speech and signal …, 2018 | 5 | 2018 |
Explaining off-policy actor-critic from a bias-variance perspective TH Fan, PJ Ramadge arXiv preprint arXiv:2110.02421, 2021 | 4 | 2021 |
Transformer Working Memory Enables Regular Language Reasoning and Natural Language Length Extrapolation TC Chi, TH Fan, AI Rudnicky, PJ Ramadge Findings of the Association for Computational Linguistics: EMNLP 2023, 2023 | 3 | 2023 |
Latent Positional Information is in the Self-Attention Variance of Transformer Language Models Without Positional Embeddings TC Chi, TH Fan, LW Chen, AI Rudnicky, PJ Ramadge Proceedings of the 61st Annual Meeting of the Association for Computational …, 2023 | 1 | 2023 |
A Contraction Approach to Model-based Reinforcement Learning TH Fan, P Ramadge International Conference on Artificial Intelligence and Statistics, 325-333, 2021 | 1 | 2021 |
A New Social Network Model of Online Forums TH Fan, KC Chen GLOBECOM 2017-2017 IEEE Global Communications Conference, 1-6, 2017 | 1 | 2017 |
Attention Alignment and Flexible Positional Embeddings Improve Transformer Length Extrapolation TC Chi, TH Fan, AI Rudnicky Findings of NAACL 2024, 2023 | | 2023 |
Advancing Regular Language Reasoning in Linear Recurrent Neural Networks TH Fan, TC Chi, AI Rudnicky NAACL 2024, 2023 | | 2023 |
System and method for controlling large scale power distribution systems using reinforcement learning TH Fan, Y Wang, U Muenz US Patent App. 17/814,535, 2023 | | 2023 |
Principles and Applications of Discrete Deep Generative Models TH Fan Princeton University, 2023 | | 2023 |
Safety Control for Prime Focus Spectrograph TH Fan, AR Kumar, PJ Ramadge 2022 56th Annual Conference on Information Sciences and Systems (CISS), 269-274, 2022 | | 2022 |